During a recent exhibition on Antarctica, our attention was drawn to spectacular structures, visible in some icebergs (one can find on the Internet views of such structures in icebergs of various origins). These are made of sets of vertical prisms of ice, or parallel planes (also vertical), forming colonnades. These last ones are separated by horizontal surfaces from the rest of the iceberg above (the tilts of these ones modify these directions); the diameters of the prisms and the thicknesses of the plans appear decimetric to pluri-decimetric. The structures are inscribed in the mass of the ice, the marine erosion and the melting highlighting them (under overhangs of the massive iceberg). To our knowledge, there does not seem to be any reasoned description, nor proposal of explanation. We are content here to paint a potential interpretative framework. We do not give in detail the reasons that guide us; they are inspired by our knowledge of the solidification of geological magmas giving comparable structures: columns and planes (see references below). We imagine that when it enters the sea, a piece of glacier, which constitutes an important "cold reserve", can, under certain conditions, make the sea freeze on its contact in its immersed parts (there is thus formation of new ice). As the sea is rich in NaCl, the solidification point of the water depends on the salt content in the vicinity of the ice/sea interface. Depending on the temperature gradients involved, the phenomenon of supercooling may come into play, and the ice/water interface shows an instability; this is responsible for the formation of digitations, i.e. fingers, or parallel planes of ice, protruding into the water during freezing. Materials scientists discuss the three-dimensional temperature gradients that allow a bifurcation between the columnar and planar regimes. Because of their growth, the fingers/planes will weld together; they will leave joints between them corresponding to more fragile zones, highlighted by the later erosion/ melting processes, during the iceberg uplift. If water ice cannot contain salt, it cannot contain a whole series of chemical elements, including gases, which will be "expelled" from the ice. Is this an explanation for the gaseous releases observed along the joints by divers (Raphaël Sané, pers. com., 2022) and sometimes considered as responsible for the structures (whereas they would be a consequence)? We will see in the future if this framework should be kept as it is, amended, or abandoned. It invites us to make new observations, measurements of various types (chemical, petrofabric), to read about related subjects, to debate with competent people. Preliminary estimates involving the main parameters of the problem (temperatures, enthalpy of fusion, heat capacities, thermal and molecular diffusivities, thermal gradients, slope of the liquidus in the phase diagram, intervention of solutal convection) are encouraging. We thank our colleagues glaciologists of Toulouse and Grenoble for first exchanges (among them Frédérique Rémy, Christian Vincent), as well as, in a special way, Raphaël Sané (naturalist guide) for his photographs and his useful remarks (without wanting to compromise any in the presented hypotheses). We salute the mediators of the Rotonde (CCSTI de la Loire and Ecole des Mines) for the exhibition (fall 2021) at the origin of this reflection. Guy B. Guy B., J. Volc. Geoth. Res., 2010; Guy B., Thiéry V., Garcia D., Bascou J. & Broekmans M., Miner. Petrol, 2020.
Bernard Guy (Sat,) studied this question.